67 C h a p t e r 4

VISUALIZING PATHOGENIC PROTEIN SYNTHESIS DURING INFECTION

4.1 Abstract

We extend cell-selective bioorthogonal noncanonical amino acid tagging (BONCAT) to

visualize staphylococcal protein synthesis in three dimensions within skin abscesses. We

use BONCAT methodology combined with the hybridization chain reaction (HCR) to

visualize both proteins and rRNA within cleared abscesses. We hypothesize that this

methodology can be readily applied to diverse microbial systems to study the

biogeography of host-microbe interactions.

This work was a collaboration with Will H. DePas, Bryan Yoo, Sarkis K. Mazmanian, and Dianne K. Newman.

68 4.2 Introduction

The biogeography, or spatial arrangement of microbes within a host, depends on the

features of their microenvironment such as nutrient availability, oxygen levels, and host-

cell interactions. Characterizing these spatial arrangements is important to understanding

and controlling microbes as they interact with the host (1, 2). While thin sectioning of

abscesses has led to remarkable insights into host-microbe interactions, tools to study the

three-dimensional biogeography of infection are currently limited. Using tissue-clearing

techniques to visualize infection may produce insights previously unobtainable by thin

slicing alone.

Recent advances in tissue clearing techniques such as the passive CLARITY technique

(PACT) have been used to render tissues transparent and allow for visualization of fine

structure in three-dimensions (3). Importantly, PACT preserves the spatial relationships of

molecules within the cell by encasing them within an acrylamide-based hydrogel, and is

compatible with most immunohistochemistry or in situ hybridization techniques (4-6).

While most researchers using PACT have focused on neuroscience applications due to the

method’s ability to preserve neural structure within the brain, the Tobin lab has used

PACT to visualize granulomas of Mycobacterium tuberculosis expressing a fluorescent

protein in clarified murine lung tissue and whole zebrafish (7, 8).

Recently, the Newman lab developed the microbial identification after passive CLARITY

technique (MiPACT) to study polymicrobial infections within sputum samples from

69 patients with cystic fibrosis (CF) (9). This method uses hybridization chain reaction

(HCR) to identify and visualize microbial species by labeling their ribosomal RNA

(rRNA). HCR uses short DNA probes complementary to a target RNA to trigger the

polymerization of fluorophore-labeled DNA hairpins in a hybridization chain reaction,

effectively amplifying signal from in situ hybridization (10, 11). By using HCR and

PACT, they were able to both quantify growth rates of CF pathogens and visualize

substructures of multi-community biofilms in sputum.

We hypothesized that cell-selective bioorthogonal noncanonical amino acid tagging

(BONCAT) could be used with tissue-clearing to explore both the protein localization and

substructures formed by Staphylococcus aureus during skin infection. In this chapter, we

apply MiPACT to staphylococcal abscesses that have been cell-selectively labeled using

BONCAT to visualize both pathogenic proteins and rRNA during infection.

4.3 Results

Staphylococcal abscesses can be clarified using PACT. As detailed in Chapter 3, we

first labeled proteins synthesized by methicillin-resistant Staphylococcus aureus (MRSA)

within murine skin abscesses using azidonorleucine (Anl) (Fig 4.1A). After excision and

fixation of the abscess, we tested various ratios of bisacrylamide, acrylamide, and

paraformaldehyde (PFA) for embedding conditions to encase the skin abscess within a

stabilizing hydrogel. The conditions we settled on were 4% of a 29:1 acrylamide:bis-

acrylamide (v/v) mixture, 1% paraformaldehyde (PFA), and 0.25% of the thermal-

initiator VA-044 in 1X PBS, which we called “B4P1”.

70

Figure 4.1: A) Scheme used to label skin abscess during infection, embed within hydrogel, and

remove lipids. B) Pictures of the abscess during incubation in SDS and after submersion in

refractive index matching solution (RIMS). Each square represents 1 mm2.

Other conditions tested, with smaller quantities of bisacrylamide or PFA were less robust

to the subsequent protocols: the inner staphylococcal abscesses would often separate from

the epidermal layer. After polymerization, we cut the abscess into ~1 mm thick slices and

removed the lipids from the tissue using the passive clarity technique (PACT). We took

pictures of these skin abscess pieces over 15 days to monitor clearing (Fig 4.1B). We did

not notice additional clearing after day 12, and on day 15 we resuspended the tissue in

refractive index matching solution (RIMS) for 24 hours. The skin abscess appeared

71 significantly more transparent (Fig 4.1B). Collagenase treatment prior to clearing steps

did not show enhanced transparency.

BONCAT/HCR labeling is compatible with PACT. To determine optimal click

chemistry and HCR conditions to use on the skin abscesses, we first tested several

conditions on B4P1 hydrogel blocks of S. aureus +MetRS or –MetRS strains grown in

tryptic soy broth (TSB) and labeled with azidonorleucine (Anl) for 1 hour during mid-

exponential growth. We tested the copper-catalyzed alkyne-azide cycloaddition (CuAAC)

with a terminal alkyne-functionalized tetramethylrhodamine (TAMRA) or the strain-

promoted alkyne-azide cycloaddition (SPAAC) using aza-dibenzocyclooctyne

functionalized with TAMRA, before and after processing the samples with HCR probes.

We found that the only conditions that showed signal from both the click reaction and

HCR was when we performed the click reaction first using SPAAC, followed by a

washing step, then incubation with HCR probes (Fig S4.1, Fig 4.2A).

We then took the azide-tagged and clarified skin abscess pieces, performed SPAAC with

DBCO-TAMRA, then performed HCR with 30 nM staphylococcal-specific initiator probe

(STA3) and corresponding amplification hairpins tagged with AlexaFluor488 (120 nM,

B4-AF488). After washing unbound probes away followed by 24 hours of incubation in

RIMS, we visualized abscesses and saw signal for both ribosomal RNA and newly-

synthesized proteins within the abscess (Fig 4.2C). Probes with a “mismatch” sequence

(no complementarity to target RNA) showed little background signal in the AF488

72 channel (Fig S4.2A) and samples infected with a –MetRS strain showed no signal in the

TAMRA channel (Fig 4.2B), suggesting that the signal was specific. First attempts to stain

skin abscesses showed a high degree of background labeling in the thick epidermal layer,

but this signal was decreased by adding 1 µM of a random DNA primer to the

hybridization step, effectively blocking charged regions that nonspecifically bind the DNA

probes within the tissue.

Figure 4.2: A) Scheme used to conjugate DBCO-fluorophore to newly synthesized proteins,

followed by HCR for S. aureus ribosomal RNA (rRNA). B) Fluorescence microscopy of skin

abscess pieces (volume ~30 mm3) after following steps described in Fig 4.1A and 4.2A. Scale bars

are 100 µm. C) Staining for azide-tagged proteins (red) overlaid in three-dimensional space with

HCR for staphylococcal rRNA.

73 Notably, the colocalization of the HCR signal and the BONCAT signal was

inconsistent: though the two signals did appear in the same general area of the abscess,

their spatial arrangement and patterns were markedly different. This implied that the

newly synthesized azide-tagged staphylococcal proteins were not in the same location as

the rRNA used to synthesize them. In several experiments the rRNA probes appeared to

surround the staphylococcal proteins in a fibrous “halo”, while in others the HCR signal

appeared offset from the bulk of the MRSA abscess as visualized through BONCAT and

DAPI. (Fig 4.2C). This result was puzzling, as we had expected the HCR signal to appear

as puncta colocalized with protein visualization. We repeated the experiment but swapped

the fluorophore used so that SPAAC was performed using DBCO-AF488 and the HCR

amplification hairpins were conjugated to AlexaFluor647 (AF647). We again saw the

HCR signal for staphylococcal rRNA slightly offset from the newly-synthesized

staphylococcal proteins visualized with BONCAT (Fig 4.3).

74

Figure 4.3: Wider field view of skin abscess following a similar protocol to Fig 4.2 except with

different fluorophores.Scale bars are 100 µm.

Comparison of BONCAT/HCR labeling using PACT with thin sections. We aimed to

compare the biogeography observed with PACT to that observed with traditional thin

sectioning techniques. We cryosectioned fixed Anl-labeled staphylococcal abscesses to

~15 µm thick before applying BONCAT and HCR probes. Similar to the PACT results,

we observed that the HCR signal colocalized only partially with the BONCAT signal.

Furthermore, in the cryosectioned samples there was a strong BONCAT signal showing

newly-synthesized staphylococcal proteins in a deeper layer of the skin abscess that we

had previously not observed in the PACT samples (Fig 4.4). This BONCAT signal

colocalized perfectly with polyclonal anti-S. aureus immunostaining, confirming the

identity of staphylococcal proteins. We also compared these microscopy images to

hematoxylin and eosin-stained (H&E) paraffin slices (~10 µm), in which the abscess

75 demonstrates a dark purple color due to the large amount of polymorphonuclear

leukocytes (neutrophils, or PMNs) (Fig 4.4).

Figure 4.4: Comparisons of various techniques to visualize staphylococcal skin abscesses. Scale

bars are 100 µm. In the samples cleared by PACT (~3x3x3 mm in volume), the HCR signal

surrounded the BONCAT signal. In the cryosections, a similar pattern was seen to the PACT

samples in the top portion of the skin (towards the epidermis), but an additional strong BONCAT

signal (DBCO-488) was seen in the lower layers of the skin (towards the muscle). Additionally,

the BONCAT signal for newly-synthesized proteins overlays with an anti-S. aureus polyclonal

antibody. After paraffin embedding, thin-sectioning, and H & E staining, the neutrophils (dark

purple) can be seen in several layers of the abscess within the skin.

76 4.4 Discussion and Future Directions

The colocalization of the anti- S. aureus immunostaining with signal from BONCAT

suggests we have successfully tagged newly-synthesized staphylococcal proteins within

the skin. HCR results are less clear as they do not exactly colocalize with signal from S.

aureus proteins within this skin abscess infection model.

We have several testable hypotheses as to why part of the abscess lacked HCR signal: 1)

the rRNA in this area is being degraded prior to detection, 2) the BONCAT and

immunostaining is visualizing lysed S. aureus within host phagocytes, 3) the BONCAT

and immunostaining is visualizing secreted proteins from S. aureus. During skin infection,

S. aureus secretes numerous toxins to lyse host cells, sometimes while within phagocytes

(12, 13), and BONCAT has previously been shown to tag secreted proteins (14).

Chatterjee et al. showed that ~60% by mass of the proteins secreted by a S. aureus in vitro

culture was a class of toxins known as phenol soluble modulins (PSMs) (15). Further

immunohistochemistry experiments using antibodies against these toxins could confirm

the identity of potential proteins secreted into the abscess.

Additional RNA preservation steps (RNase inhibitors, for example) may be required for

these hydrogel-embedded tissues. However, because the HCR signal was similar in both

cleared tissues and cryoslices, we reasoned that penetration of the DNA probes in the

abscess is not the issue, even within the dense hydrogel network. The strong BONCAT

signal colocalized with anti-S. aureus immunostaining in the area not labeled by HCR was

only seen in the cryoslices, and not in the clarified skin tissue. Future experiments could

77 elucidate whether this is because the hydrogel-embedded tissue is losing this part of the

abscess, or if the clearing step is removing some newly-synthesized tagged proteins.

Increased amounts of bisacrylamide and PFA may lend increased stiffness to the skin

layers to preserve the intact abscess, but decreasing the hydrogel porosity may reduce

probe penetration and increase artifacts such as swelling.

While tissue-clearing processes such as PACT have not reported on the loss of large

amounts of proteins before, it is known that fluorescent lipophilic molecules like DyeI

(MW: ~1 kDa) are incompatible with these techniques as they associate with the lipids

removed from cell membranes. Several secreted staphylococcal toxins such as PSMs and

the related delta-toxin are small (2-3 kDa), detergent-like cytolytic proteins that associate

with host cell membranes and form transient pores within the membrane (16). It is

possible that the tissue clearing detergents used in PACT remove this class of proteins,

potentially revealing a limitation to this method in the visualization of pathogenic

proteins in host tissue. We could test this hypothesis by assaying the PACT solution

post clearing for proteins using Western blotting or LC-MS/MS (4).

We have shown the chemistries used for conjugation of azide-tagged molecules are

compatible with the HCR and PACT techniques pioneered by MiPACT. We envision cell-

selective labeling using BONCAT could be combined with tissue-clearing techniques to

not only visualize host-pathogen interactions, but neural proteomic mapping as well.

78 4.5 Supplementary Information

Supplemental Figure S4.1 DBCO-TAMRA labeling of incorporated azides and HCR staining for

S. aureus in B4P1 hydrogels. Only the +MetRS samples show signal in the DBCO-TAMRA

channel, while both the –MetRS and +MetRS samples show HCR signal for staphylococcal RNA.

Scale bars are 50 µm.

79

Supplemental Figure S4.2: Controls for HCR in A) PACT abscesses and B) cryoslices.

Mismatch-RNA probes show no binding to the abscess and staphylococcal probes do not bind to

uninfected skin.

80

Supplemental Figure S4.3 A) BONCAT signal overlays with α-S. aureus in cryoslices. Only

samples with +MetRS show signal in the DBCO-AF488 channel, while both +MetRS and MetRS

samples stain positively for S. aureus proteins using an anti-S. aureus antibody. B) Staphylococcal

rRNA signal does not entirely colocalize with BONCAT signal.

Experimental Procedures

Strains and growth conditions. The following strains were used in this study:

Staphylococcus aureus USA300, JE2 (NARSA), grown in tryptic soy broth (TSB)

aerobically with shaking at 250 rpm.

Embedding labeled bacteria grown in vitro in hydrogel blocks. We used similar

procedures as DePas et al with slight modifications. (9). S. aureus strains –MetRS

(pWW412, see Appendix A) or +MetRS (pSS20_hprk, see Appendix A) were inoculated

from single colonies from tryptic soy agar (TSA) plates in TSB with chloramphenicol (20

81 μg/mL). When the cultures reached mid-exponential phase, 2 mM of azidonorleucine

(Anl) was added for 1 hour. The cultures were washed with PBS, and fixed overnight at 4

°C in 4% paraformaldehyde (PFA). Fixed samples were resuspended in 29:1

acrylamide:bis-acrylamide (v/v) (Bio-Rad 161-0146) and 0.25% VA-044 hardener (w/v)

(Wako 27776-21-2) in 1X PBS for polymerization. After removing oxygen from the

solution in an anaerobic chamber, blocks were polymerized at 37 ºC for three hours,

without shaking, and then cut to ~3 mm3. The staphylococcal cells were then digested with

lysostaphin (50 μg/mL) in 50 mM Tris buffer for 2 hours at 37 °C. Samples were washed

twice in PBS, then “cleared” for 5 days in 8% SDS in PBS at 37 °C.

Ethics statement. Animal experiments were performed in accordance with the regulations

for the Institutional Animal Care and Use Committee (IACUC) at Caltech.

Embedding murine skin abscesses in hydrogel blocks. S. aureus strains -MetRS or

+MetRS were prepared as described in Chapter 3 for skin infection. After labeling with

Anl for 16 hours, the mice were culled (CO2) and the abscesses excised. They were

directly added to 4% paraformaldehyde (PFA) for fixation overnight at 4 °C then washed

with PBS. Fixed samples were resuspended in 29:1 acrylamide:bis-acrylamide (v/v) (Bio-

Rad 161-0146) 1% paraformaldehyde (PFA, EMS #15713), and 0.25% VA-044 hardener

(w/v) (Wako 27776-21-2) in 1X PBS for polymerization. The staphylococcal cell wall

was then digested with lysostaphin (50 μg/mL) in 50 mM Tris buffer for 6-12 hours at 37

82 °C. Samples were washed once in PBS, then cleared for 5-21 days in 8% SDS in PBS at

37 °C.

Click Reaction. After washing the SDS out of the cleared samples with three washes of

PBS, free cysteines were blocked with 100 mM iodoacetamide in PBS at room

temperature for 4-16 hours. Oftentimes, samples are reduced prior to blocking, but using

10 mM dithiothreitol (DTT) before iodoacetamide treatment showed higher background

labeling. DBCO-TAMRA or DBCO-AF488 (Click Chemistry Tools) was added to a fresh

solution of 100 mM iodoacetamide to a concentration of 5 μM and the reaction proceeded

for 1 hour. The samples were washed twice in PBS (30 min each), rotated end over end in

50% DMSO in PBS overnight, then washed three more times in PBS.

HCR. 5’-ATTTCACATTTACAGACCTCAACCTACCTCCAACTCTCAC-3’ was

added to the 3’ end of the DNA probe (termed “B4” (10)). DNA hairpins (Molecular

Instruments) conjugated to either AlexaFluor-488 or AlexaFluor-647 as indicated were

used with the appropriate initiator probe sets.

We initially used a previously-reported STA3 probe (17), but found high signal-to-noise

with this sequence, so we increased the length of our probe from 16 nucleotides to 31

nucleotides. Microscopy images of the HCR in abscesses were all performed with this

STA3_long probe. Sequence in Supplementary Table 1.

83 Hybridization: Samples were hybridized in 500 μL of HCR hybridization buffer (100

μL of 20X SSC, 100 mg dextran sulfate (Sigma D6001), 250 μL formamide, ddH2O to 1

mL) with 30 nM initiator probe at 46 ºC, with shaking, for 24-48 hours. All solutions were

filter sterilized. Excess probe was removed by washing each sample in 50 mL 84 mM

FISH wash buffer (840 μL of 5 M NaCl, 1 mL of Tris-HCl [pH 7.6], 500 μL of 0.5 M

EDTA [pH 7.2], 100 μL of 5% SDS, and Milli-Q H2O to 50 mL) at 52 °C for 6 hours in a

water bath.

Amplification: Hairpin pairs were first heated at 95 °C for 90 seconds in a thermocycler in

separate PCR tubes, then cooled at room temperature for 30 minutes. Each hairpin in a

pair was added to a final concentration of 115 nM. Amplification buffer with the

appropriate hairpin mixture (120 μL) was then added to each sample in a 1.5 mL

centrifuge tube. Samples were incubated at room temperature with gentle shaking for 24-

48 hours. After amplification, samples were washed in 337.5 mM FISH wash buffer

(3375 μL of 5 M NaCl, 1 mL of Tris-HCl [pH 7.6], 500 μL of 0.5 M EDTA [pH 7.2],

100 μL of 5% SDS, and Milli-Q H2O to 50 mL) at 48 °C for 3 hours in a water bath.

Samples were then incubated in 250 μL RIMS with 10 μg/mL DAPI (1:1000 from 10

mg/mL stocks in DMSO) at room temperature for at least 16 hours before imaging.

Microscopy. Prior to imaging, samples were incubated at RT, with shaking, overnight in

RIMS with 1 μg/mL DAPI. Samples were then mounted on slides in 0.9 mm or 1.7 mm

Coverwell perfusion chambers (Electron Microscopy Services) with a coverslip on the

top. Imaging was performed using a Zeiss LSM 880 confocal microscope with a Plan-

84 Apochromat 10x/0.45 M27 objective (WD 2.0 mm). All images and Z-stacks were

collected in 12-bit mode, with at least 1024x1024 scan format. Images were processed

using Imaris imaging software (Bitplane) or the FIJI distribution of ImageJ (18).

Histological examination of mouse skin abscesses. Mouse skin was harvested 2 days

after inoculation and fixed in 10% neutral-buffered formalin for 48 hours. Fixed tissues

were embedded in paraffin, sectioned (5 μm), mounted on slides, and stained with

hematoxylin and eosin (Pacific Pathology).

Cryosectioning, DBCO-staining, immunostaining, and imaging. Abscesses were

excised, placed in optimum cutting temperature (O.C.T.) compound, and frozen at -80 °C.

They were then cut into ~15 μm sections and deposited onto glass slides. The slides were

fixed with 2% paraformaldehyde (PFA), permeabilized with 0.1% Triton-X, and treated

with lysostaphin at 37 °C. The samples were next blocked in 100 mM iodoacetamide for

30 min in the dark, then reacted with 5 μM DBCO-AlexaFluor488 (Click Chemistry

Tools) for 15 min. After washing, the samples were blocked with 1% mouse serum, then

treated with a 1:3000 dilution of anti-Staphylococcus aureus antibody (polyclonal, rabbit)

(Ab37644, Abcam). A goat anti-rabbit secondary antibody conjugated to AlexaFluor555

(Thermo Scientific) was then added at 1:10,000 dilution. Samples were washed with PBS,

and Vectashield Antifade with DAPI was added prior to coverslips. Imaging was

performed using a Zeiss LSM 880 confocal microscope with a Plan-Apochromat 10/0.45-

85 numerical aperture M27 objective (WD 2.0 mm) or a 25X objective. Image

reconstruction and analysis was performed in the FIJI distribution of ImageJ (18).

Supplemental Table 1: Probes used in this study:

Species Name Sequence Staphylococcus STA3 (17) GCACATCAGCGTCAGT Staphylococcus STA3_long GATCCCCACGCTTTCGCACATCAGCGTCAGT

- Mismatch ACTCCTACGGGAGGCAGC - Random AGCAGGTCGAACTCCTTGAG

References

1. Stacy A, McNally L, Darch SE, Brown SP, & Whiteley M (2016) The

biogeography of polymicrobial infection. Nat Rev Microbiol 14(2):93-105. 2. Jorth P, et al. (2015) Regional Isolation Drives Bacterial Diversification within

Cystic Fibrosis Lungs. Cell Host Microbe 18(3):307-319. 3. Lai HM, Ng WL, Gentleman SM, & Wu W (2017) Chemical Probes for

Visualizing Intact Animal and Human Brain Tissue. Cell Chem Biol 24(6):659-672.

4. Chung K, et al. (2013) Structural and molecular interrogation of intact biological systems. Nature 497(7449):332-7.

5. Yang B, et al. (2014) Single-Cell Phenotyping within Transparent Intact Tissue through Whole-Body Clearing. Cell 158(4):945-958.

6. Shah S, et al. (2016) Single-molecule RNA detection at depth by hybridization chain reaction and tissue hydrogel embedding and clearing. Development 143(15):2862-2867.

7. Cronan MR, et al. (2016) Macrophage Epithelial Reprogramming Underlies Mycobacterial Granuloma Formation and Promotes Infection. Immunity 45(4):861-876.

8. Cronan MR, et al. (2015) CLARITY and PACT-based imaging of adult zebrafish and mouse for whole-animal analysis of infections. Dis Model Mech 8(12):1643-1650.

9. DePas WH, et al. (2016) Exposing the Three-Dimensional Biogeography and Metabolic States of Pathogens in Cystic Fibrosis Sputum via Hydrogel Embedding, Clearing, and rRNA Labeling. Mbio 7(5): e00796-16.

10. Choi HMT, Beck VA, & Pierce NA (2014) Next-Generation in Situ Hybridization Chain Reaction: Higher Gain, Lower Cost, Greater Durability. ACS Nano 8(5):4284-4294.

11. Dirks RM & Pierce NA (2004) Triggered amplification by hybridization chain reaction. Proc Natl Acad Sci 101(43):15275-15278.

86 12. Thurlow LR, et al. (2013) Functional Modularity of the Arginine Catabolic

Mobile Element Contributes to the Success of USA300 Methicillin-Resistant Staphylococcus aureus. Cell Host Microbe 13(1):100-107.

13. Li M, et al. (2010) Comparative analysis of virulence and toxin expression of global community-associated methicillin-resistant Staphylococcus aureus strains. J Infect Dis 202(12):1866-1876.

14. Mahdavi A, et al. (2014) Identification of secreted bacterial proteins by noncanonical amino acid tagging. Proc Natl Acad Sci 111(1):433-438.

15. Chatterjee SS, et al. (2013) Essential Staphylococcus aureus toxin export system. Nat Med 19(3):364-367.

16. Otto M (2014) Staphylococcus aureus toxins. Curr Opin Microbiol 17:32-37. 17. Tavares A, Inacio J, Melo-Cristino J, & Couto I (2008) Use of fluorescence in

situ hybridization for rapid identification of staphylococci in blood culture samples collected in a Portuguese hospital. J Clin Microbiol 46(9):3097-3100.

18. Schindelin J, et al. (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9(7):676-682.